OPINION

The Association of Irritated Residents, El Comité para el Bienestar de Earlimart, the Community of Children’s Advocates Against Pesticide Poisoning, and the Natural Resources Defense Council, petition for review of a final action by the Environmental Protection Agency approving in part and disapproving in part revisions to California’s State Implementation Plan for meeting air quality standards for ozone under the Clean Air Act. We have jurisdiction under 42 U.S.C. § 7607 (b)(1). We grant the petition for review and remand to EPA for further consideration.

I

A

Congress enacted the Clean Air Act (the “Act”) to help protect and enhance the nation’s air quality. 42 U.S.C. §§ 7401 -7671q. The Act requires the Environmental Protection Agency (“EPA”) to establish National Ambient Air Quality Standards (“NAAQS”) for a variety of pollutants, one of which is ozone. Id. §§ 7408-09. EPA then designates areas as “attainment” or “nonattainment” based on whether the areas meet the clean air standards for each particular pollutant. Id. § 7407(d). EPA classifies nonattainment areas based on the severity of the area’s pollution, from Marginal to Extreme. Id. § 7511(a). The area at issue in this litigation — the Los Angeles-South Coast Air Basin (“South Coast”) — is classified as Extreme. 73 Fed.Reg. 63,408, 63,409 (Oct. 24, 2008) (to be codified at 40 C.F.R. pt. 52).

Under the Act, states have primary responsibility for ensuring that the quality of their air satisfies the NAAQS, and they must detail their efforts in a State Implementation Plan (“SIP”) for each region within that state. 42 U.S.C. § 7410 (a). States must submit these SIPs and SIP revisions to EPA for review. EPA may either fully approve the plan, partially approve and partially disapprove the plan, or conditionally approve the plan. Id. § 7410(k). Once approved, SIPs become enforceable as federal law. Id. § 7413.

An EPA determination that a state has failed to submit a required plan, or EPA disapproval of a submitted plan, triggers two time periods. First, a “sanctions clock” begins during which time the state must either remedy the deficiency or face sanctions. Id. § 7509(a)-(b). Second, a “FIP clock” begins by the end of which EPA must either approve a state-submitted SIP or promulgate a Federal Implementation Plan (“FIP”). Id. § 7410(c)(1). Additionally, EPA must issue a “SIP call,” and thereby require the state to make necessary revisions, if it finds that a previously approved SIP is “substantially inadequate” to attain or maintain air quality standards. Id. § 7410(k)(5).

The Act also contains “conformity” requirements. Under these conformity provisions, the federal government may not approve, accept, or fund any transportation plan, program, or project unless it conforms to an approved SIP. Id. § 7506(c). To make conformity determinations, transportation agencies must look to an approved SIP to find the maximum amount of pollution allowed from motor vehicle emissions. This motor vehicle emissions budget (“MVEB”) is determined by the states in their SIPs by identifying the total allowable emissions consistent with meeting the statutory clean air requirement, and then allocating that total among various types of sources, such as motor vehicles. 40 C.F.R. § 93.101 . Because SIPs sometimes take years to review, EPA may make preliminary adequacy determinations regarding the MVEBs found in the SIPs. Id. § 93.118. After further review, EPA may declare the MVEB to be inadequate. Id. § 93.118(e)(3).

In addition to the SIPs and conformity requirements applicable to all areas, the Act contains further requirements for nonattainment areas, depending on the severity of the ozone problem in the area. 42 U.S.C. §§ 7511 -7511f. Two of these requirements are at issue in this case. The first requirement is for these nonattainment areas to submit SIP revisions demonstrating attainment of the ozone standard by the applicable date. These “attainment plans” have two main parts: (1) a control strategy to reach compliance; and (2) an attainment demonstration to show that under the strategy the area will meet the NAAQS by the statutory deadline. Id. §§ 7511a(c)(2)(a), (d)-(e); 7410(a)(2)(A).

The second requirement for nonattainment areas is to develop enforceable transportation strategies and control measures “to offset any growth in emissions from growth in vehicle miles traveled ... and to attain reduction in motor vehicle emissions as necessary.” Id. § 7511a(d)(1)(A). Suggested transportation control measures include programs for improved public transit, restrictions of certain lanes for high occupancy vehicles, and programs for secure bicycle storage facilities. Id. § 7408(f)(1)(A).

B

In 1994, California submitted a SIP revision that included an ozone attainment demonstration for the South Coast nonattainment area and a “Pesticide Element” designed to reduce emissions from pesticide applications. In 1997, EPA approved the SIP revision with respect to both the ozone attainment demonstration and the Pesticide Element. In 1999, California sought again to update the SIP with new emissions inventories and a new ozone attainment demonstration. EPA approved these elements in 2000. All of these plans and revisions form the 1997/1999 South Coast Ozone SIP (“1997/1999 SIP”).

After EPA approved the 1997/1999 SIP, California conducted new modeling, demonstrating that the existing SIP underestimated vehicle pollution in the area. Specifically, California realized that, with respect to ozone:

Concluding that “a plan update [was] necessary,” California submitted the 2003 SIP Revision to EPA in 2004. The 2003 SIP Revision consisted, in relevant part, of three things: the 2003 Attainment Plan, PEST-1, and a demonstration that no transportation control measures were required. Petitioners seek review of EPA’s final determination as to each of these three elements.

The 2003 Attainment Plan revised the existing SIP in two ways. First, it updated the attainment demonstration (and therefore the MVEBs) to account for the increased emissions projections under the new modeling. Second, it added additional control measures to offset the increase in predicted pollution. In 2004, EPA found the MVEBs in the attainment demonstration adequate for purposes of the conformity provisions, but did not make a final decision as to the 2003 Attainment Plan as a whole. In 2008, California withdrew some of the 2003 Attainment Plan’s key elements, including many of the control measures.

PEST-1 was a control strategy that called for continued implementation of the Pesticide Element approved in the 1997/1999 SIP. In 2008, we concluded that the portion of the Pesticide Element representing the enforceable commitment (the Wells Memorandum), was not part of the 1997/1999 SIP. See El Comité Para El Bienestar de Earlimart v. Warmerdam, 539 F.3d 1062 (9th Cir.2008).

Finally, because nonattainment areas must propose enforceable transportation control measures “to offset any growth in emissions from growth in vehicle miles traveled,” California submitted to EPA a demonstration purporting to show that there would be no such growth in emissions. Although California acknowledged that vehicle miles traveled would increase by about 30%, it showed that aggregate motor vehicle emissions would decrease.

In 2008, EPA proposed to approve the control measures that were not withdrawn from the 2003 Attainment Plan (including PEST-1), but disapprove the attainment demonstration (and therefore the MVEBs) in the 2003 Attainment Plan because the demonstration was largely based on the withdrawn commitments. EPA explained the consequences of its proposed partial disapproval by saying: “No sanctions clocks or FIP requirement would be triggered by our disapprovals, if finalized, because the approved [1997/1999] SIP already contains the plan elements that we are proposing to disapprove.” 73 Fed. Reg. 63,408 , 63,419 (Oct. 24, 2008) (to be codified at 40 C.F.R. pt. 52). EPA also proposed to approve California’s assertion that no transportation control measures were required based on California’s demonstration that there would be no growth in aggregate vehicle emissions.

In 2009, after considering public comments from petitioners on the proposed rule, EPA finalized action on the 2003 SIP Revision as proposed. This timely petition for review followed.

Petitioners raise three issues in their petition for review. First, they contend EPA’s failure to order California to submit a revised attainment plan for the South Coast after it disapproved the 2003 Attainment Plan was arbitrary and capricious. Second, petitioners contend EPA’s approval of PEST-1 violates the Clean Air Act because PEST-1 lacks enforceable commitments. Third, petitioners contend EPA violated the Act by failing to require transportation control measures to combat the increase in vehicle miles traveled. We grant the petition as to all three claims.

II

EPA’s failure to evaluate the adequacy of the existing SIP was arbitrary and capricious. The Act requires each nonattainment area to submit a SIP that includes an attainment demonstration for the 1-hour ozone standard. 42 U.S.C. § 7511a(c)(2)(A). Although EPA approved such an attainment demonstration for the South Coast in the 1997/1999 SIP, California submitted a revised attainment demonstration in the 2003 Attainment Plan, which made clear that the attainment demonstration in the 1997/1999 SIP was not accurate. When EPA partially disapproved the 2003 Plan’s attainment demonstration (because California subsequently revoked many of the control strategies on which the attainment demonstration was based), EPA concluded that no further action was required. See 74 Fed.Reg. 10,-176, 10,177 (Mar. 10, 2009) (to be codified at 40 C.F.R. pt. 52) (“[N]o sanctions clocks or Federal Implementation plan (FIP) requirement[ are] triggered by our disapprovals because the plan revisions that are the subject of the proposed disapprovals represent revisions to previously-approved SIP elements that EPA determined met the CAA requirements, and thus, the revisions are not required under the Act.”).

EPA is mistaken that its duties under the Act end upon approval. Instead, EPA had an affirmative duty to evaluate the existing SIP and determine whether a new attainment demonstration was necessary to ensure California satisfies the Act’s attainment requirements. Its failure to evaluate the adequacy of the existing SIP in any way was arbitrary and capricious.

Through the 2003 SIP Revision, EPA knew, or should have known, of the inadequacy of the 1997/1999 SIP. As California specifically stated, “this revision points to the urgent need for additional emission reductions (beyond those incorporated in the 1997/99 Plan) to offset increased emission estimates from mobile sources and meet all federal criteria pollutant standards within the time frames allowed under the federal Clean Air Act.” EPA’s public comments also indicate that it understood that the new modeling undermined the existing SIP. See, e.g., 73 Fed.Reg. 63,408, 63,415 (Oct. 24, 2008) (to be codified at 40 C.F.R. pt. 52) (“[I]n view of the magnitude of the reductions now understood to be needed for attainment of the 1-hour ozone NAAQS in the South Coast, [California] has adopted [additional control measures].”); id. at 63,416 (“[California] revised the 1-hour ozone attainment demonstration in the 2003 South Coast AQMP in light of updated emissions inventories that show higher mobile source emissions than prior projections and updated modeling that indicates a lower carrying capacity in the air basin.”). However, even if EPA did not actually know the extent to which the new modeling undermined the existing SIP, it has a duty to evaluate the adequacy of the existing SIP as a whole when approving SIP revisions. See Hall v. EPA, 273 F.3d 1146, 1159 (9th Cir.2001) (“The EPA must be able to determine that, with the revisions in place, the whole ‘plan as ... revised’ can meet the Act’s attainment requirements.”) (quoting Train v. Natural Res. Def. Council, 421 U.S. 60, 90 , 95 S.Ct. 1470 , 43 L.Ed.2d 731 (1975)). In partially approving the 2003 Plan, EPA should have analyzed the adequacy of the whole 1997/1999 SIP.

The closer question is whether, given the knowledge that a previously approved SIP likely no longer meets the Act’s attainment requirements, EPA has an affirmative obligation to request a new attainment demonstration. Two sections of the Act may give rise to such an affirmative duty. The first is the requirement that EPA issue a FIP when it disapproves any plan or plan revision. 42 U.S.C. § 7410 (c)(1); see also id. § 7509 (mandating sanctions upon disapproving a required State submission). The second is the requirement that EPA issue a SIP call upon a finding that the existing SIP is substantially inadequate. Id. § 7410(k)(5).

EPA argues that the FIP and sanction clocks are triggered only where the plan revision is “required” under the Act. T he first step in statutory construction cases is to begin with the language of the statute. Barnhart v. Sigmon Coal Co., Inc., 534 U.S. 438, 450 , 122 S.Ct. 941 , 151 L.Ed.2d 908 (2002). The plain text refutes EPA’s argument. Section 7410(c)(1) states:

42 U.S.C. § 7410 (c)(1). Although subsection (A) applies to “required” submissions, subsection (B), which applies to disapprovals of SIPs and SIP revisions, does not have such a limit. “Nor should we infer as much, as it is a general principle of statutory construct that when ‘Congress includes particular language in one section of a statute but omits it in another section of the same Act, it is generally presumed that Congress acts intentionally and purposely in the disparate inclusion or exclusion.’ ” Barnhart, 534 U.S. at 452 , 122 S.Ct. 941 (quoting Russello v. United States, 464 U.S. 16, 23 , 104 S.Ct. 296 , 78 L.Ed.2d 17 (1983)).

EPA contends that reading subsection (B) as requiring EPA to issue a FIP whenever it disapproves a discretionary revision would “yield absurd results, because it would require the agency to promulgate a FIP where the State’s fully approved SIP remains in effect.” EPA provides as an example a state proposing a revision to make an existing SIP less stringent, arguing that requiring a FIP when EPA disapproves such a relaxing of the SIP would be irrational. EPA’s point is well taken in a situation where the existing SIP remains adequate to attain the NAAQS. The facts in this case, however, are much different because EPA knew, or should have known, that the “fully approved SIP” was no longer adequate. While it may seem counterintuitive to require EPA to promulgate a FIP when it disapproves a revision seeking to undercut an effective existing SIP, it is entirely logical to require EPA to promulgate a FIP when it disapproves a revision seeking to update what it recognizes are serious deficiencies in an existing SIP.

Furthermore, although the plain language requires a FIP every time EPA disapproves a plan revision, the FIP can be avoided if an existing plan is in place that meets the Act’s requirements because § 7410(c)(1) has a grace period in which states can bring their plans into compliance before the FIP is enacted. See 42 U.S.C. § 7410 (c)(1) (mandating a FIP “unless the State corrects the deficiency”); see also id. § 7509(a) (mandating sanctions “unless such deficiency has been corrected within 18 months after the finding, disapproval, or determination”). EPA must simply evaluate the existing SIP (as already required under Hall, 273 F.3d at 1159 ), and if it meets the Act’s requirements, EPA can find that the state has “corrected the deficiency,” 42 U.S.C. § 7410 (c)(1).

This analysis aligns with Congress’s intent in writing the statutory language to mandate promulgation of a FIP upon any disapproval. In 1988, when Congress was debating the amendments to the Clean Air Act, EPA sought an amendment that would have left promulgation of FIPs solely to EPA’s discretion. See Coal, for Clean Air v. S. Cal. Edison Co., 971 F.2d 219, 223 (9th Cir.1992) (citing S. 1630, 101st Cong., § 105 (1989)). Although the Senate passed such an amendment, a House Committee deleted the language, and the “House language retaining EPA’s mandatory obligation to promulgate a FIP whenever it disapproves a SIP was ultimately enacted by Congress and signed into law.” Id. (citing Clean Air Act Amendments of 1990, Pub.L. No. 101-549, 104 Stat. 2399).

Even if, as EPA argues, the FIP requirement is triggered only when the revision was “required” under the Act — and not upon every EPA disapproval of a plan revision as we determined above — the partial disapproval of the 2003 SIP Revision here still triggers the FIP requirement because large portions of the 2003 Attainment Plan were not discretionary. For example, the Act explicitly requires states with nonattainment areas to update their SIPs every three years with a revised inventory of actual emissions from all sources of relevant pollutants. See 42 U.S.C. §§ 7502 (c)(3); 7511a(1)(3)(A). Additionally, the Act requires transportation projects to conform to the existing SIP and states that, “[t]he determination of conformity shall be based on the most recent estimates of emissions, and such estimates shall be determined from the most recent population, employment, travel and congestion estimates....” Id. § 7506(c)(1)(B). The conformity provisions thereby require a state to submit a SIP revision to ensure the MVEBs in the SIP are current, otherwise the state will not be able to receive federal funding. Indeed, the 2003 SIP Revision explicitly referenced these two reasons in explaining why it submitted the changes:

In summary, EPA’s duty to issue a FIP (or sanctions) represents one statutory source of EPA’s duty to take further action upon partial disapproval of California’s 2003 Attainment Plan.

Alternatively, EPA’s obligation to take further action can be derived from the statutory requirement that the Administrator issue a SIP call upon a finding that the existing SIP is substantially inadequate. 42 U.S.C. § 7410 (k)(5) (“Whenever the Administrator finds that the applicable implementation plan for any area is substantially inadequate to attain or maintain the relevant national ambient air quality standard ... or to otherwise comply with any requirement of this chapter, the Administrator shall require the State to revise the plan as necessary to correct such inadequacies.”). EPA argues that the decision about when and whether to review an existing SIP for substantial inadequacy is entirely within the Administrator’s discretion. In support of this argument, EPA cites the text of the statute and two out-of-jurisdiction cases. The text of § 7410(k)(5), however, only says that the Administrator must make the finding, not that the finding must be a product of Administrator-initiated review procedures. The two cited cases also provide little support for EPA because they only show that the Administrator must have some discretion in deciding whether to find a SIP substantially inadequate. See Sierra Club v. Johnson, 541 F.3d 1257, 1265-66 (11th Cir.2008); Citizens Against Ruining the Env’t v. EPA 535 F.3d 670, 677-78 (7th Cir.2008). We do not dispute this point. However, the question is not whether EPA has discretion in determining substantial inadequacies exist, but whether EPA has unlimited discretion to ignore evidence indicating an existing SIP might be substantially inadequate and choose to do nothing. We believe EPA’s failure to act in light of the strong evidence provided in the 2003 SIP Revision demonstrating the substantial inadequacies of the 1997/1999 Plan is arbitrary and capricious. See Motor Vehicle Mfrs. Ass’n of U.S. v. State Farm Mut. Auto., 463 U.S. 29, 43 , 103 S.Ct. 2856 , 77 L.Ed.2d 443 (1983) (“Normally, an agency rule would be arbitrary and capricious if the agency ... entirely failed to consider an important aspect of the problem....”); see also 1000 Friends of Maryland v. Browner, 265 F.3d 216, 235 (4th Cir.2001) (leaving open the possibility that “there may be cases where previously performed modeling is inadequate to demonstrate attainment such that EPA’s failure to require new modeling in those cases might be found to be arbitrary or capricious”). EPA’s decision to do nothing is especially troublesome in light of the Act’s overall purpose of ensuring states come into corn pliance with clean air standards. See 42 U.S.C. § 7470 .

EPA also notes that a demonstration that the 1997/1999 SIP is outdated or ineffective is not equivalent to finding that the SIP as a whole is substantially inadequate because a SIP is a complex, multi-faceted set of obligations. Again, the determination about whether the SIP is substantially inadequate is within the Administrator’s discretion. We merely determine that the Act requires EPA to evaluate the existing SIP and actually make the determination as to whether a new attainment demonstration is required.

Because EPA’s failure to evaluate the adequacy of the existing SIP was arbitrary and capricious in light of the 2003 SIP Revisions alerting EPA to the new modeling, we grant the petition for review. Specifically, EPA has an affirmative duty to ensure that California demonstrate attainment with the NAAQS, see 42 U.S.C. §§ 7410 (a)(2)(A), 7502(c)(6), either by promulgating a FIP, issuing sanctions, or evaluating the necessity of a SIP call.

Ill

EPA’s action in approving the pesticide element of the SIP was arbitrary and capricious. EPA approved PEST-1— the portion of the 2003 SIP Revision recommitting to implementing the Pesticide Elements from the 1997/1999 Plan — in its 2009 final action. Petitioners claim our decision in Warmerdam, 539 F.3d at 1072 , in which we stated the Wells Memorandum was not part of the existing SIP, rendered the Pesticide Element’s commitments discretionary, thereby violating the Act. See 42 U.S.C. § 7410 (a)(2)(A) (“Each implementation plan ... shall include enforceable emission limitations and other control measures, means, or techniques ... as well as schedules and timetables for compliance.”); § 7502(c)(6) (same).

EPA does not address the merits of this contention. It only argues that petitioners lack standing to challenge the Pesticide Element. Specifically, EPA argues that petitioners’ injuries were not caused by EPA’s 2009 rulemaking and cannot be redressed by the relief they seek. We disagree.

Petitioners bear the burden of demonstrating a causal connection between their injuries and EPA’s conduct. See Lujan v. Defenders of Wildlife, 504 U.S. 555, 560 , 112 S.Ct. 2130 , 119 L.Ed.2d 351 (1992). EPA argues that because its 2009 action approving PEST-1 merely maintained the status quo with respect to the Pesticide Element, either approval or disapproval would have resulted in the same regulatory outcome: continuation of the existing Pesticide Element as approved by EPA in 1997. EPA’s argument assumes incorrectly that approving PEST-1 does not require an evaluation of the existing Pesticide Element as part of the SIP as a whole. As we determined above, when EPA approves a plan revision, it must ensure that the whole plan, as revised, satisfies the Act’s requirements. Hall, 273 F.3d at 1159 . This responsibility is even more important where, as here, the revision simply reiterates the commitments of the prior plan.

EPA also claims it had a “false choice” because its approval of PEST-1 did not make the Pesticide Element any more or less enforceable. This contention is not entirely true. Although EPA approved an identical plan in the 1997/1999 SIP, it wasn’t until our 2008 Warmerdam decision that EPA approved the plan with the knowledge that the plan may not include enforceable commitments. As first submitted in 1994, EPA worried the Pesticide Element did not meet the requirements of the Act, primarily because it failed to in- elude specific dates for adoption and implementation of the regulations necessary to achieve the required reductions. See Warmerdam, 539 F.3d at 1067 . EPA did not propose approval of the Pesticide Element until California submitted the Wells Memorandum, which committed to adopting any necessary regulations by specific years and in specific areas. Id. In proposing approval of the Pesticide Element, EPA responded to questions about the Pesticide Element’s enforceability by citing to the Wells Memorandum, thereby indicating its belief that the Wells Memorandum provided the required enforceable commitments. See, e.g., 62 Fed.Reg. 1150, 1169-70 (Jan. 8,1997) (to be codified at 40 C.F.R. pt. 52). After Warmerdam , EPA affirmatively knew that the Wells Memorandum was not part of the 1997/1999 SIP, and therefore its approval of PEST-1 (and by incorporation the existing Pesticide Element) in light of this knowledge represents the causal link giving rise to petitioners’ injuries.

EPA further argues petitioners cannot demonstrate redressability because if it disapproves PEST-1 on remand, the existing Pesticide Element as approved in 1997 would remain in effect. As we determined above, however, any disapproval of a SIP revision triggers the FIP and sanction clocks unless EPA determines the existing Pesticide Element has sufficiently enforceable commitments to meet the Act’s requirements. See 42 U.S.C. §§ 7410 (c)(1); 7509. Therefore, a remand is required to allow EPA to make that determination.

IV

EPA’s failure to require transportation control measures was arbitrary and capricious. Petitioners contend EPA violated the Act when it partially approved the 2003 SIP Revision without requiring California to submit transportation control measures to offset the emissions resulting from an increase in vehicle miles traveled. EPA argues that because aggregate motor vehicle emissions will decrease each year, California did not need to adopt control measures. The disagreement centers on one sentence in the Act requiring transportation control measures “to offset any growth in emissions from growth in vehicle miles traveled.” 42 U.S.C. § 7511a(d)(1)(A). The relevant sentence states in full:

Id. § 7511a(d)(1)(A). Petitioners argue that in determining whether to impose TCMs, EPA should identify the level of emissions emanating solely from VMT in a prior year, and use that as the baseline from which to measure the change in emissions. EPA’s current approach, in contrast, is to use the aggregate emissions from a prior year as the baseline against which to measure the change in emissions. Aggregate motor vehicle emissions reflects the combination of numerous variables unrelated to VMTs such as vehicle turnover, tailpipe control standards, and use of alternative fuels. Because the parties agree that VMTs will increase by around 30%, but that aggregate motor vehicle emissions will decrease, the question for the court is whether “any growth in emissions” can mean any growth in aggregate motor vehicle emissions, or is unambiguous in meaning any increase in the level of emissions solely from VMTs.

To interpret § 7511a(d)(1)(A), we utilize Chevron’s “familiar two-step procedure.” Nat’l Cable & Telecomm. Ass’n v. Brand X Internet Servs., 545 U.S. 967, 986 , 125 S.Ct. 2688 , 162 L.Ed.2d 820 (2005); Chevron, U.S.A., Inc. v. Natural Res. Def. Council, 467 U.S. 837 , 104 S.Ct. 2778 , 81 L.Ed.2d 694 (1984). To determine whether the phrase “to offset any growth in emissions from growth in [VMT]” is ambiguous, we must determine “whether Congress has directly spoken to the precise question at issue.” Chevron, 467 U.S. at 842 , 104 S.Ct. 2778 . “If the intent of Congress is clear, that is the end of the matter; for the court, as well as the agency, must give effect to the unambiguously expressed meaning of Congress.” Id. at 842-43 , 104 S.Ct. 2778 . At Chevron step one, if, employing the “traditional tools of statutory construction,” we determine that Congress has directly and unambiguously spoken to the precise question at issue, then the “unambiguously expressed intent of Congress” controls. Id. at 843 , 104 S.Ct. 2778 . In determining congressional intent, we not only examine the precise statutory section in question but also analyze the provision in the context of the governing statute as a whole, presuming a congressional intent to create a “symmetrical and coherent regulatory scheme.” FDA v. Brown & Williamson Tobacco Corp., 529 U.S. 120, 131-33 , 120 S.Ct. 1291 , 146 L.Ed.2d 121 (2000). We also examine legislative history. See N. Cal. River Watch v. Wilcox, 620 F.3d 1075 , 1083 (9th Cir.2010).

We begin with the plain words of the statute. The use of the word “growth” in reference to both “emissions” and “vehicle miles traveled” suggests two baselines: one pegged to changes in emissions and the other pegged to changes in VMT. EPA argues that petitioners’ interpretation reads the phrase “growth in emissions” out of the statute because the Act would then only require a simplistic analysis of whether VMT is increasing. Although EPA is correct in stating that any increase in VMT is very likely to result in an increase in aggregate emissions, we cannot ignore the possibility that with advances in clean car technology, one day VMT could increase without a corresponding increase in emissions. If that happens, under the statute, EPA would not need to impose TCMs even though VMT increased. Therefore, although some increase in emissions is required (such that there are two baselines), it doesn’t change the ultimate question of whether the baseline for the increase in emissions can be viewed in terms of aggregate vehicle emissions (as EPA contends), or if the baseline must be viewed as any increase in emissions due solely to VMT.

EPA’s interpretation only gives effect to the second clause of the relevant sentence, and not to the first. According to the statute, states shall implement TCMs not only “to offset any growth in emissions from growth in [VMT]” but also “to attain reduction in motor vehicle emissions as necessary ... to comply with the ... periodic emissions reduction requirements[ ].” 42 U.S.C. § 7511a(d)(1)(A). While the second clause contemplates using TCMs to reduce aggregate emissions, the first clause contemplates using TCMs to reduce VMT. See United States v. Wenner, 351 F.3d 969, 975 (9th Cir.2003) (utilizing principles of statutory construction to determine that general, catchall provisions should not trump more specific provisions). Looking at both clauses not only demonstrates that EPA’s interpretation — equat ing “growth in emissions” with “growth in aggregate emissions” — is redundant, it shows that Congress used the phrase “motor vehicle emissions” when referring to aggregate emissions, but simply “emissions from growth in [VMT]” when referring to only those emissions from VMT.

This interpretation is bolstered by the legislative history of the provision, which clearly refutes EPA’s interpretation. The House Committee Report, for example, specifically states how “growth in emissions” should be measured, explaining: “The baseline for determining whether there has been growth in emissions due to increased VMT is the level of vehicle emissions that would occur if VMT held constant in the area.” H.R. REP. NO. 101— 490, pt. 1, at 242 (1990). This Report is very persuasive because, “the authoritative source for finding the Legislature’s intent lies in the Committee Reports on the bill, which ‘represent] the considered and collective understanding of those Congressmen involved in drafting and studying proposed legislation.’ ” Garcia v. United States, 469 U.S. 70, 76 , 105 S.Ct. 479 , 83 L.Ed.2d 472 (1984) (quoting Zuber v. Allen, 396 U.S. 168, 186 , 90 S.Ct. 314 , 24 L.Ed.2d 345 (1969)). EPA even admits that “it is true that the language of [the House Committee Report] appears to support the alternative interpretation of the statutory language,” and that “the original authors of the provision and [the House Committee Report] may in fact have intended this result.” 57 Fed.Reg. 13,498, 13,522 (April 16, 1992) (to be codified at 40 C.F.R. pt. 52).

Further review of the legislative history provides additional support for our conclusion. See, e.g., S. Rep. NO. 101-228, at 44 (1989) (“Severe and extreme areas are required to offset growth in vehicle miles traveled by implementing the transportation controls listed....”); 136 Cong. Rec. 16,956 (1990) (statement of Sen. Max Baucus) (“It is clear that the goals of this bill — a healthy and safe air supply for every American — will not be achieved without implementing strategies that effectively limit the growth in vehicle use in the major urban centers where pollution levels are the worst.”). Therefore, because the statutory language and the legislative history demonstrates that Congress has spoken directly to the question at issue, we do not owe deference to EPA’s interpretation, Chevron, 467 U.S. at 842-43 , 104 S.Ct. 2778 ; Wilderness Society v. Fish & Wildlife Service, 353 F.3d 1051, 1062 (2003), and we grant the petition for review.

V

In summary, EPA’s approval of the 2003 SIP Revision was arbitrary and capricious. EPA should have ordered California to submit a revised attainment plan for the South Coast after it disapproved the 2003 Attainment Plan. EPA should have required transportation control measures. EPA is required to determine whether the Pesticide Element has sufficient enforcement mechanisms to satisfy the requirements of the Act. We grant the petition for review and remand to the EPA for further proceedings consistent with this opinion.

PETITION GRANTED.

. Ground-level ozone is a primary component of what is commonly known as "smog.” It is formed when oxides of nitrogen (NOx), volatile organic compounds (VOC), and oxygen react in the presence of sunlight, generally at elevated temperatures. When inhaled, even at very low levels, ozone can cause serious health problems by damaging lung tissue and sensitizing lungs to other irritants. See 73 Fed.Reg. 63,408, 63,409 (Oct. 24, 2008) (to be codified at 40 C.F.R. pt. 52). . Although the 2003 SIP Revision references the triennial requirement in the California Clean Air Act, the Federal Clean Air Act also mandates such a triennial inventory. See 42 U.S.C. §§ 7502 (c)(3); 7511a(1)(3)(A).